Home Made Fuel Injector Cleaning Machine

[60dgrzbelow0]

In lieu of using glass or Pyrex... these are identical to the ones I'm using on "The FrankInjector Machine" for Flow Balance Tests and are 50ml graduated polypropylene cylinders...

http://cgi.ebay.com/LOT-24-MEASURING-CYLINDERS-50ML-Lab-Laboratory-Plastic-/350265489665?cmd=ViewItem&pt=LH_DefaultDomain_0&hash=item518d733901

Those EFI connectors are the lowest priced ones I've seen yet... Great Find!

[HOYS]

After reading the tech sheets for the LMD 18400, his hand written diagrams, and pictures he took, I'm pretty sure I know what he is doing with the second controller board.

[ollopa]

After reading the tech sheets for the LMD 18400, his hand written diagrams, and pictures he took, I'm pretty sure I know what he is doing with the second controller board.

You can see in the pictures that he brings all four inputs to the driver out to a screw terminal but only has the one white wire connected. The other screw terminals must be jumpered to the single input from the 555+LM393 board. Not much of anything going on with the driver board. It's plug and play.

The way he adds a comparator to the 555 in order to control pulse-width is interesting. He has a circuit that's basically like this one to set the frequency:


But he takes the output from the top of the capacitor at pin 2 of the 555 instead of the square wave at pin 3. He calls the waveform a sawtooth but it actually resembles a distorted triangle wave (exponential rise and exponential fall whereas a sawtooth would have a linear rise and an abrupt vertical drop). This he feeds into one input of a comparator in the LM393. The other input would be connected to the wiper of a potentiometer that just acts as a voltage divider and sets the pulse-width. Like this:


The output of that comparator then goes to all four inputs of the LMD18400. But for about half as much as that part costs, you could duplicate the following circuit for each injector and tie the inputs in parallel. Use a 12V supply to power the comparator so its outputs are high enough to fully turn on the MOSFETs:


I'd suggest a 10 to 100 Ohm resistor between each MOSFET gate and comparator output, though, to reduce parasitic oscillations. See how each parallel MOSFET is connected here (just look at the MOSFETs and ignore the left and right ends of the circuit):


There's a pulldown to ground to ensure they aren't turned on by static electricity and to discharge the gate capacitance so they don't get stuck on, and then each one is driven through a 10 Ohm resistor.

Combine all of these circuits and calculate the potentiometer and capacitor values for the frequency range you want for the 555.

[60dgrzbelow0]

Jesus Palomino! What I would not give to possess a clone of just one lobe from your brain that understands all of this. This is pretty clever detective work, Rick. Okay... so now that you have broken down all of the performance values into discreet circuits... is this anything like you intend to eventually build? If not... would the average radio Shack repair technician be able to hold the hand of an idiot like me and show us how to make one that works from off-the shelf parts? This is fairly exciting stuff to me...even though I don't really understand any of it from looking at your circuit images. I just can't WAIT to see what comes of this!

[ollopa]

Naw, it's really quite basic stuff to an electronics guy. Double entry accounting and ERP are a mystery to me, but I'm sure your typical MBA could talk my ear off all day in business jargon.

I don't think Radio Shack still has repair technicians. None of the employees at the last 5 Radio Shacks I visited knew the difference between a resistor and a diode. They're just pushy cell phone salespeople now. In typical form, one girl insisted on helping me find a part and then asked me "how do you spell 'resistor?'" I doubt any of the people there are going to be able to help you. They probably couldn't even point you to the right books that would help you.

But I think Hoys can fill in the gaps I left. Try an IRF510 and 1N4001 for the driver. Heat sink the IRF510 and you can probably drive 4 injectors off the same MOSFET.

What I'm putting together has a continuous mode where it's free-running and you can turn a knob to control the frequency. I did away with variable duty-cycle, though, and fixed it at 50%. I've yet to see any utility in varying the duty-cycle for this application. It's the pulse-width that matters. Anyway, it can be reconfigured in software if we later discover a need for it.

Still waiting for my cylinders to come in. I never got a tracking number or shipping confirmation and my inquiry to the company has gone unanswered. Assuming they were shipped, I'm expecting arrival some time this week. I just painted some rusty headers. I don't have a sandblaster so I used electrolysis to clean the rust. I'd recommend giving that method a try to anyone with rusted parts. Check it out over here: http://www.instructables.com/id/Elec...val-aka-Magic/ I got the best results with an overnight soak in the bucket. The rust mostly flakes off. What's left turns to a soft black residue that can be rubbed off with soapy water and a scotch brite pad. Kinda fits with our theme of mixing electricity and chemistry to clean car parts

What I wouldn't give for a 48-hour day or the ability to survive comfortably on only 2 hours sleep....

[60dgrzbelow0]

Well... I guess we'll all be "Standing By to Stand-By" until you fabricate your first "Magic Box" and finish your own version of "Frank" to test the thing out.

FWIW... I have an album on my photobucket site that shows my use of FiberGlas Header Wrap and also some White VHT Paint for reconditioning serviceable headers. The only problem with removing the rust and scale electrolytically is that sometimes... so much metal has rusted and flaked away that the remaining down tubes metal can wind up just about paper thin and burn through. Here is the link to the subject matter with special attention of the VHT Paint being on Page 2:

http://s557.photobucket.com/albums/ss19/60dgrzbelow0/HEADERS/?start=0

[ollopa]

Well I think it's about time to check in with a status report. My wife had a vacation last month so we went out of town for a while and had a really great time. The electrolytic rust removal and VHT paint on the headers came out great and has been holding up beautifully.

I wanted to take this opportunity to try out some different CAD tools and technologies, so I've been a little bit distracted from the injector driver board layout. I've been messing around with toner transfer PCB creation using a modified laminator and CuCl etchant. I got really good results down to 5mils! That's just amazing for at-home PCB fabrication. I've also started working on a tabletop CNC mill/drill.

Getting back to this project, I started doing the board layout in KiCAD. It's probably the best there is that's open-source and free, but it's more cumbersome to use than commercial CAD software. I don't know how many revisions of the board I'll go through before I'm satisfied with the layout, but I'm planning to make my first board in the next few days.

I ran into a problem with my injector cleaner design. I bought glass cylinders instead of plastic for some reason and the glass bases are just a little bit too big to allow me to line the cylinders up beneath the injectors. I need to trim a little material off the sides of each glass base and I'm not sure what the best method is. Right now I'm thinking about using a soldering gun to make a hot wire and pressing the wire against the glass, then submerging the glass in some cold water. From what I've read, the glass should break along the line where I applied heat. I guess if that actually works, I'll try to make a video and share.

[60dgrzbelow0]

I have not heard a word from "ollopa" since last September, but since I received an odd question via my email project address of "at60dgrzbelow0@yahoo.com" concerning something I failed to adequately explain, I'll mention it now. When I was designing the clear Acrylic piping for the first version of "Frank", I had to be concerned with the possibility of the hose kinking and complete collapse whenever any acute angles or bends happened along the length of the Acrylic tubing. This might seem like a trivial problem, but keeping that tubing perpetually open is a critical problem with the machine and its solution came to me from an unusual set of experiences.

I remembered seeing something during many homicide autopsies I had to witness involving deaths by manual strangulation. It was common for the Forensic Pathologist to entirely remove the larynx and upper bronchial airway of the victim in order to dissect and get better view whether or not the victim's Hyoid Bone was crushed... a sure sign of non-accidental death from suffocation. When viewing these body parts up close, you can see the intimate details of that area, noting now that in our own anatomy, there are cartilaginous rings externally surrounding the windpipe at precise flexing intervals in the neck and upper body to prevent the collapse of the airway, should the windpipe be compromised by any external pressure or atmospheric changes while inhaling or exhaling.

Unlike the design of a vacuum hose, with ribs and indentations along the inside of the tube way, having something there to brace the spherical airway leaving the tubular interior surfaces glistening and glass smooth, thus NOT inducing any convolutions or places for junk and liquids to collect and obstruct breathing is important to long term health. It follows that I had to do something similar with the Acrylic tubing in order to ensure that all of the EFI Cleaning Solution would make its way down inside the clear plastic gallon jug acting as the collection container. It occurred to me to use thin, strong plastic nylon straps/ties that when cinched up tightly at critical points along the outside of the tubing, the presence of those ties would force the tubing to remain open, even at points of greatest curvature or bending. And so to answer the person's question at last... The reason for the black nylon ties...is quite simply to allow "Frank"...to breathe... under the worst of circumstances...

[Tbay99Venture]

The Frankinjector thread lives! I wondered why there wasn't an update in a while, but then I have been 60*V6 deprived for a while since I have been swamped with work, kids & honey-do things.

Still, I'm wondering what bits and pieces on the electrical end are left hanging. It looked like a pretty straightforward circuit posted a while back. Since it appears the low-impedance injectors aren't used a specialty controller chip isn't required. (That's what tripped me up when I first started looking around, but it's been old news to me since May.) Is this snag now just the wish to duplicate exactly what the PCM is doing, rather than some grouping of them in banks that the power supply could tolerate?

[60dgrzbelow0]

Welcome back, Brother... I have been in the same boat as far as other projects and demands... Life somehow refuses to remain linear and thus... we all have to bounce around from one project to another. The short answer about whether the emulation of all the features of control exhibited by the PCM is necessary is ... No. Since it is not critical to have the injectors alternate their firing order in a normal ignition timing pattern and approaching a level of stoichiometric perfection is unnecessary, There really are only three things that need to be done on "Frank" to provide the means necessary in controlling how it works and achieve the completion of a proper cleaning and a set of performance tests to proof out sets of any EFIs to sanitary and fully functioning conditions:

1. The EFIs must be cycled at rates of 50 ms. 100ms and 1 second intervals to ensure that the internal metering valve and spring are up to snuff...and perform at rates for each speed/duration evenly, one with each other. If cycled like this again and again (at the moment, the only reasonably priced actuators are those from OTC), you will be able to determine if any injectors indicate problems electromagnetically... If they are sluggish and non-responsive after having their filter basket(s) replaced and after a thorough cleaning in an ultra-sonic hot tank and a through fluid pressure cleaning via "The FrankInjector machine"... then check them with a multi-meter for defective innards

2. The EFIs must be able to fire all at once in order to visualize the 50mL cylinders and observe the rise of the EFI cleaning fluid levels for comparison against each other as a set. The GM factory recommendation for the Rochester Multec lines of EFIs indicates they should all be within 5% of each other to prevent imbalances between cylinders through a wide range of metering needs...say from a 600 RPM idle all the way up through the entire operational range of the engine(s) performance.

3. The "ECM_Lite" version of the electronic controller must be straightforward and robust enough to function via leads from a 12 VDC Car Battery.

Here is a short but informative EFI sales video that prompts the end consumer to avoid cleaning their EFIs and suggests they clean and/or buy all new ones at 15,000 mile intervals. But the information in this video provides a lot of important factual data worth looking at whether the viewer is an old hand at this project...or a newcomer curious to know why this post has been "hit" almost 15,000 times since we started this unusual, but wholly successful repair project:

http://www.youtube.com/watch?v=ueJkjLgVJVk

[robertisaar]

i'm still very serious about using audio files to control the injectors...

just have to figure out what combinations of components will allow me to combine a standard 1/8" stereo(or mono) cable from a laptop(or boombox, walkman, MP3 player, what have you, anything that can play CDs or just audio files) and using something like Audacity to create files with different lengths of silence/noise so that the injectors only fire when there is "sound" happening... unfortunately, i'm not that savvy with building my circuits around this stuff yet, though i imagine it requiring a solid-state relay(home-made if necessary) and a transistor or two... some pretty low-buck stuff IMO.

and being able to go beyond Bob's stated 50/100/1000mS on cycles will actually allow you to determine/adjust appropriate values in tunes to compensate for injector opening times based on supply voltage and minimum pulsewidths required to get consistent action.

plus, i'd kinda like to know what the super mario brother's theme sounds like with injectors.

[pocket-rocket]

plus, i'd kinda like to know what the super mario brother's theme sounds like with injectors.

That would be epic, lol.

[60dgrzbelow0]

Hello Robert... Long time ... No Hear... Glad you are Okay...

I know this doesn't make sense...But... instead of using sound... (and since Ollopa has disappeared from "site" (LOL) with his "Gizmo-in-Tow" ) What if I continue to work on "Frank" by using a child's pinwheel inside of a shoebox and after hooking up a photoelectric relay on one end... put a flashlightat the other end and turn a small fan on the pinwheel...and watch "Frank" Rock n' Roll while spritzing the hell out of a set of EFIs..."

Pocket... See the attached image of the thing Max mentioned to you....

[robertisaar]

i've actually seen a solid state relay made using a LED and phototransistor, so it's certainly workable.

and the disappearing act was unintentional Bob.

[60dgrzbelow0]

It wasn't you that "disappeared", Robert......but Ollapa.... I feel like Augustus Caesar...Yelling at his troops for not getting back his "Eagles" when his troops were wiped out in Gaul...after the Germans butchered his legions and then stole his standards. In this case... Frank needs the thing Ollapa said he would build to "complete me"... I really want this project completed... I have so many other ideas I want to get to get work on... Like "The FrankenUberLuber Machine" which I think is the next thing really worth building...